Introduction to the Techniques of Paninian Grammar- Chetan Pandey Bhagawan (Lord) Panini – as he is reverentially called by Sanskrit Pundits – like most names found in Ancient Indian History is of uncertain date. We have some evidence that he was born in what us today the North West Frontier Province of Mode rn Pakistan in a th village called Shalatur . Scholars have surmised the 5 Century B.C.E as a possible period of his existence. But he could have been earlier.
Lord Panini is one of the greatest geniuses to have emerged from the subcontinent of India who had a formative and sustaining influence on Indian Culture and Civilization. He ranks amongst Patanjali , Nagarjuna and Tulsidas , men who made an original and revolutionary contribution to World Civilization. It is no exaggeration to say that Indian Civilization rests on the shoulders of Panini . In a short grammatical treatise called the Ashtadhyayi Ashtadhyayi – or the Book in Eight Chapters -, he codified all the rules that govern the concurrent spoken Sanskrit of his times. Thus he made the Sanskrit Language systematic and free from ambiguity. This gave the Sanskrit Language a precise structure and no place for variant spellings or forms making it a meticulous tool for higher philosophical speculations and other kinds of scientific and scholarly work. With such a standardized language, much like Modern English, interregional communication and textual interpretation became unequivocal and free from vagueness. It is no wonder that Hindu India continued using Sanskrit well into the end of Nineteenth Century as the Official Language of its Courts and Religious Institutions Institutions because of its well-organized structure. His grammatical treatise has been rightfully called ‘one of the greatest monuments of human intelligence’ by the American Linguist Leonard Bloomfield (quoted in ‘Panini : A Survey of Research, George Cardona, pg 243).
Need for a Sanskrit Grammar There is much truth in the saying that ‘Necessity is the Mother of Invention’ and in the case of Sanskrit Grammar and Linguistics nothing is truer. But wha t was the necessity? Being of little pragmatic value, not one of the great ancient civilizations of ancient ancien t China, Middle East, Ancient Egypt, Europe( with the partial exception of Greece ) have pondered upon the linguistics principles which govern human speech; but Panini’s g rammar, phonology, semantics, morphology Ashtadhyayi Ashtadhyayi is a mine of information on grammar, etc. The question arises why Panini and other scholars before him were trying to understand human speech, sound patterns, compounding, tenses, sandhi (euphonic combination), word roots etc. The answer lies in the Vedas – a set of three, later four – tomes which were revered as the word of GOD.
The ritual called ‘Yagna ’ was the most important religious practice of the ancient Aryans and was central to the communities’ spiritual and intellectual life. Just as techno logy and science are the chief focus of modern world, Yagna was the chief focus of the ancient Aryans. This ritual, which is as much part of modern Hindu religion as millennia before, consists of a fire Altar over which Vedic hymns are chanted to honor the sectarian deities like Agni (the Fire God), Indra ( God of Material Prosperity and Rain ), Saraswati( Goddess of Learning ) etc. It was believed that even a single syllable along with the tones - which wh ich disappeared later and are not seen in any modern Indian language including the Prakrits - ill-pronounced would gain the wrath of the deities and lead to rebirth in Hell. Making matters worse, writing down the Vedas was prohibited as was transmitting it to non-Brahmins. The Vedic meters called Mantras are said to loose their efficacy if they are written down and must be received orally only from a wise Brahmin Guru. Mantras pronounced in the prescribed manner unerringly can alter the shape of the Universe, upset the laws of Karma and bestow b estow spiritual and material blessings like longevity of life and wisdom to the officiating o fficiating priest and his client. Besides destroying enemies and curing illnesses. A well-known parable in the Mahabhashya commentary relates how the demon Vritra mispronounced a single tone in the compound word ‘Indra-shatru’ (enemy of Indra) while offering oblations with the chant ‘indrashatrurvardhasva’ (may (I) - the ‘destroyer of Indra’ - prosper) thereby altering the meaning to (may ‘Indra, the destroyer’ prosper) leading to the mantra empowering Indra, the god of the Noble Deities and Vritra’s slaying by the former. Nagarjuna, called the greatest Buddhist philosopher, has described his philosophy Madhyamika as something which misunderstood can be fatal like a snake held from the wrong end or a mantra mispronounced. Contemporary practitioners and scholars like Swami Agehananda Bharati (Austria-born Indologist and Anthropologist who ordained as a Hindu Sanyasin and wrote prolifically) p rolifically) and Frits Staal (Netherlands-born Professor Emeritus, Department of South and South Asian Studies, University of California, Berkeley, Vedic Scholar par excellence) affirm that this indeed is the case with mantric praxis. To ensure that posterity does not get a corrupted version of their mantric corpus, the learned Brahmins tried to ponder on the phonological and grammatical principles which govern human speech. A pre-paninian grammatical work by Yaska called ‘Nirukta’, survives which gives us a glimpse of their proto-grammatical musings. Over time knowledge accumulated, theories refined, more discoveries made, culminating in the Grammatical system of Panini who derives many of o f his methods and terminologies from his predecessor which were doubtless many as he himself quoted around ten ancient authorities( Sphotayana, Yaska, Galava etc ).
This unprecedented work of Sanskrit Grammar and Linguistics – a fruit of a tradition of several centuries of grammatical thinking – wiped out without a trace all other parallel systems of learning grammar of which several comprehensive systems were extan t during Panini’s time and became the sole standard standa rd followed ever after. Pundits began saying ‘apaniniyam na prayunjeet’ – Do not use non-Paninian forms. There is a popular anecdote of the Weight-carrying Brahmin which shows the intolerance for non-standard usage. An old Brahmin was carrying a very heavy load on his shoulders and walking when a man sees him and asks ‘Does not that heavy weight hurt you (badhati , wrong usage though intelligible, like saying have you drunken - instead of drunk- your milk). milk). The Brahmin replied, replied, it does not hurt me( badhate , proper usage ) as much as your badhati ( wrong form) hurts me(badhate ).
Structure of the Ashtadhyayi (Book in Eight Chapters) The Book Ashtadhyayi is divided into eight chapters -hence the name. Each chapters is divided into Four Quarters (or Foot like that of mammals) containing unequal numbers of Sutras – or aphorisms, packed with linguistic and grammatical observation, theories and rules. The Entire book is in pithy aphorisms which - besides having a mnemonic value v alue - are described as a metalanguage and have the structure of algebraic formulas. Brevity is the principal feature of this work and Panini has gone to such an incredulous level in attaining it that this book is one of the marvels of aphoristic writings in the history of mankind. It is said of him that the shortening of his text by one syllable in his work gave him the pleasure of the Birth of a Son. This is in keeping with the Sutra-style of writing writing which is traditionally described in the shloka as: alpakSharamasandigdham saarvatdvishvatomukham | Astobhamanavadyam cha sutram sutravido viduh || A Sutra is one which is: (1) Expressed in few syllables (2) Free from Ambiguity (3) Comprehensive (4) Universal (5) Without inserted meaningless terms (6) Irreproachable More than two-and-a-half-millennia of scholarship has attested that Panini’s have rarely strayed from any of these aforementioned characteristics.
This treatise is unintelligible without the help of a proper Co mmentary. Patanjali, some centuries later, incorporating the scholia (Vartikas) of Katyayana, did this commendable task in his Mahabhashya. Hence, Proper Understanding of Grammar is believed to be attained by the diligent study of Mahabhashya and Grammar is also referred to as Trimuni-Vyakarana or the Grammar of the Three Sages. Contents of Ashtadhyayi The following is a topical arrangement of the chapters of Ashtadhyayi:
Book I (a) the samjna sutras(lexical items) and paribhasha sutras(meta-rules) (b) rules dealing with extension, atmane-parasmaipada and karakas(Syntax)
Book II (a) rules dealing with compounding, nominal inflections, number and gender of compounds, replacements relative to roots, ‘deletion by ‘luk’.
Book III (a) rules dealing with the derivation of roots ending in affixes ‘saN’ etc, items ending in a ‘Krit’ or ‘tiN’
Book IV-V (a) rules dealing with the derivation of a pada p ada ending in a ‘sup’, feminine affixes; derivation of nominal stems ending in an affix termed taddhita
Book VI- VII (a) rules dealing with doubling, samprasarana, samhita, augment(agama),accents, phonological operations relative to a pre-suffixial base(anga) sUT (b) rules dealing with operations relative to affixes and augments
Book VIII (a) rules dealing with dvitva(Doubling) relative to a pada, accent relative to a pada (b) rules dealing with miscellaneous operations relative to a pada and a non-pada. (adapted from Prof. Ram Nath Sharma’s ‘The Ashtadhyayi of Panini’) Panini has divided the Ashtadhyayi into two basic organizational units: the ‘sapadasaptadhyayi’ (the first seven chapters and first quarter of the Eighth) a nd the ‘tripadi’ (the remaining three quarters of the eighth). These two are blind to the effects of each other. Abbreviatory Devices Used in Ashtadhyayi Panini has used several ingenious devices to achieve economy of words and syllables, even phonemes. The foremost amongst these techniques are the Siva-Sutra-pratyahara system. The Siva Sutras
Of unknown authorship the fourteen Siva Sutra or Siva aphorisms are a catalog ca talog of phonemes divided into 14 sets. They are in essence a reordering of the traditional enumeration of the Devanagari alphabets (akshar-samamnaaya). Each of the fourteen groups contains an extraneous phoneme appended at the end technically called an ‘it ’ (meaning going away or disappearing) disappea ring) or marker and has a merely indicatory role and is not part of the set. These fourteen Siva aphorisms are:
1 2 3 4 5 6 7 8 9 10 11 12 13 14
Class of Phonemes Simple vowels Simple vowels Dipthongs Dipthongs Semi-vowels Semi-vowel Nasals Voices aspirated stops Voices aspirated stops Voices non-aspirated stops Voiceless aspirated stops Voiceless non-asp. stops spirants spirant
Member Phonemes ‘a’ ‘i’ ‘u’ ‘R’ ‘lR’ ‘e’ ‘o’ ‘ai’ ‘au’ ‘h’ ‘y’ ‘v’ ‘r’ ‘l’ ‘J’ ‘m’ ‘G’ ‘N’ ‘n’ ‘jh’ ‘bh’ ‘gh’ ‘Dh’ ‘dh’ ‘j’ ‘b’ ‘g’ ‘D’ ‘d’ ‘kh’ ‘ph’ ‘ch’ ‘Th’ ‘th’ ‘c’ ‘T’ ‘t’ ‘k’ ‘p’ ‘z’ ‘S’ ‘s’ ‘h’
‘it’ or markers ‘N’ ‘k’ ‘G’ ‘c’ ‘T’ ‘N’ ‘m’ ‘J’ ‘S’ ‘z’ ‘v’ ‘y’ ‘r’ ‘l’
Harvard Kyoto Transliteration Scheme has been used.
If Panini wants to bring up a discussion of say, dipthongs, he has two options, coin a term for ‘dipthongs’ or list them (‘e’, ‘o’, ‘ai’, ’au’). This can be cumbersome and is done away with the shorthand notation ‘ec’ which is technically called a pratyahaara pratyahaara and generated using the afore-listed Siva-Sutras . To find out what ‘ec’ signifies the first phoneme ‘e’ should be searched amongst the Column entitled ‘member phonemes’ and it is located at the beginning of Sutra number 3. The second phoneme ‘c’ should be searched amongst the Column ‘marker phonemes or ‘it’s’ and can be found in the fourth row. Thus ‘ec’ is a short-from notation for all phonemes between ‘e’ and ‘c’ (‘e’, ‘o’, ‘ai’, ‘au’)not including the markers which are merely indicatory. Similarly, ‘hl’ means all consonants, ‘jh-y’ all stops etc. In this way tens of pratyahaaras can be generated though Panini only used forty-two. Thus, Panini was saved from coining forty-two new terms and also from the awkward way of listing them. Other Abbreviatory Techniques: If a list of related affixes are introduced, the initial pho neme of the first item and the marker-phoneme of the last item in the list will mean a n abbreviation for the entire list.
For example, {‘su’, ‘au’, ‘jas’, ‘am’, ‘auT’, ‘zas’…… ‘os’, ‘sup’ } are a set of twenty-one affixes which denote case relations. ‘su’ of item one a nd ‘p’ the marker phoneme of last item ‘sup’ together generate the abbreviation ‘sup’ . Similarly ‘ting’ (‘tiJ’ in HK) is an abbreviatory term for the eighteen root-affixes {‘tip’, ‘tas’ ‘jhi’ …..’vahi’, ‘mahing’} generated by using ‘ti’ of ‘tip’ and appended to ‘ng’ of ‘mahing’. Another abbreviatory rule is given by sutra ‘aNudita savarNasya (1.1.69). Given the sets of varga-consonants, eac h set by its first-element and the phoneme ‘u’. varga-consonants, one can refer to each Hence pa-varga phonemes ‘ p’,’ph’,’b’,’bh’,’m’ p’,’ph’,’b’,’bh’,’m’ are indicated in short form by ‘ pu pu’; similarly ‘ku’, ‘chu’,’Tu’ and ’tu’. Another technique involves denoting a set of items by their shared common sounds: for example ‘Ap’ collectively for the affixes ‘TAp’, ‘chAp’, ‘ DAp’. Similarly, ‘nI’ for the affixes ‘nISh’ and ‘nIp’ and ‘nIn’ Technique of Ellipsis: Anuvrittis(recurrence) Yet another method for achieving brevity is ellipsis or stating rules only partially. The rest of the terms are supplied by words which have already occurred before. To understand ellipsis, consider the following example: a. Give John a Cow. b. A blanket to Peter c. Also to Paul Sentence (a) is complete in itself. To understand un derstand (b) we have to borrow the verb ‘give’ from the preceding context. Sentence (c) must borrow both ‘give’ and ‘blanket’ to be intelligible.
Ellipsis is technically referred to by as ‘anuvritti (recurrence)’. Consider sutra: 3.1.96 tavyat-tavya-anIyarah (the affixes ‘tavyat’, ‘tavya’, ‘anIyar’) This sutra merely mentions three affixes: ‘tavyat’, ‘tavya’, ‘anIyar’. To complete the meaning of this sutra we have to refer to preceding sutras: 3.1.1 Pratyayah (affixes) 3.1.2 Parascha (and are afterwards) 3.1.91 Dhatoh (regarding Verbal Root) Once we utilize terms from these we get the expanded meaning: The affixes (3.1.1) {‘tavyat’, ‘tavya’, ‘anIyar’ (3.1.96)} occur (introduced by convention) after (3.1.2) a Verbal Root (3.1.91)
Thus Panini has saved at least three words in one sutra. He has ha s made very extensive use of anuvritti and there are very few sutras which do not borrow terms from prior sutra. Hence, understanding ellipsis is crucial to understanding the Ashtadhya yi. Unique System of Notation Panini has used the inflections of the Sanskrit Language to indicate meaning. Consider ‘lasya’ (3.4.77) ‘tiptas….vahimahing’ (3.4.78)
3.4.77 translated into plain English will look like ‘L’s’. 3.4.78 is a listing of eighteen post-verbal affixes. These two sutras don’t make much sense. To understand 3.4.77 we have to use the paribhasha paribhasha(meta-rule) ‘shashthi sthaneyoga’ (1.1.49) A meaningless genitive implies an item ‘that will be replaced’. This can be symbolically represented as X-gen, where item X is a substituend because it has been stated in the genitive case
Hence ‘L’s’ implies that in an operation if we encounter the affix ‘L’ it must be replaced by any one of {‘tip’,’tas’….’vahi’, ‘tip’,’tas’….’vahi’, ’mahing’}. Similarly meaningless ablative endings (tasmaditi uttarasya ; 1.1.67) imply a left context: such that in a grammatical operation something applies to what follows and meaningless locative endings(tasminniti nirdishte pUrvasya ; 1.1.66) imply a right context such that in a grammatical operation something applies to what precedes.
Types of Rules Traditionally, scholars have classified the Sutras into six types. Panini has not hinted to any such classification and it is strictly Post-Paninian. However, there are obvious advantages to it and aid in the understanding of the purport of the Sutra. Modern Paniniyas have added few more types to the traditional six-fold classification. (1) sanjna sutras or technical rules: rules which assign a term for a given entity. For example in rule 1.1.1 the term vriddhi is coined meaning the set of phonemes ‘aa’, ‘ai’, ‘au’. They have been compared to the Backus-Naur Notation of Modern Computer Science. (2) paribhasha paribhasha sutras or interpretive rules: rules which provide a check on the operational rules so that they do not no t suffer from ‘over-application’, ‘underapplication’ and ‘impossible application’. For example, ‘vipratishedhe param
kaaryam’ : when two rules are equally applicable, the rules that appear later shall apply.
(3) vidhi sutra or operational rules: these form the core of the grammar. All other rules assist the operational rules. Example ‘Vartamaane lat’: To indicate the present tense, use the affix ‘lat’. (4) niyama rules or restriction rules: rules which restrict the scope of other rules. For example, Sutra ‘laH parasmaipadam’ (1.4.98) declares the replacements of ‘l’ – which are the eighteen affixes called ting - to be called parasmaidpadas. However, subsequent sutra ‘tangAnavAtmanepadam ’ (1.4.99) declares the nine ‘tang’ affixes (part of ting affixes) as Atmanepada. Therefore scope of 1.4.98 has been restricted by 14.99 and only non-tang ting affixes are to be understood as parasmaipada. (5) adhikara rules or heading rules: these rules are similar to a he ading in modern books. Adkikaras have domains which are not always well defined and only the commentaries like Kasika and Mahabhashya have to be consulted to understand their scope. Within the domain of an adhikara, all rules will pertain to it and also use parts or whole of that rule’s words to complete their meaning (refer to ‘Anuvritti’ above). For example, rule 3.1.91 is ‘dha toh’ (Verbal Roots). The word is meaningless unless understood to imply that rules till 3.4.117 are under its domain and every rule described between 3.1.91 to 3.4.117 is formulated to describe Dhatus. Also, since rules are only stated partially, the ‘phrase’ ‘Dhatoh’ has to be supplied – along with other phrases from other adhikaras which may come within the domain and extend beyond – to make these rules a complete sentence. (6) nishedh sutras or negation rules: rules which negate the provisions made by a previous Sutra. For example, rule 1.3.3 ‘halantyam ’ states that ending consonants of affixes should be treated as markers, hence hen ce disappear. But subsequent rule 1.3.4 ‘na vibhaktau tusmaah’, immediately negates this rule for nominal affixes whose consonants end with ‘t’, ‘th’, ‘d’,’dh’ ‘n’ ,’s’, ‘m’. (7) vibhasha or optional rules: Sometimes more than two rules can a pply to a given operation and hence two forms of a word are possible. This is hardly unusual, because language being a living entity, native speakers take liberties which are as common in English as Sanskrit. To account acco unt for more than two forms of a word, Panini uses Vibhasha rules to state that alternate forms are also possible. For example, rule 1.1.31 ‘vibhasha jasi’ states that the set of pronouns catalogued in the ‘sarva’ group when compounding with other words to become dvandva class of compounds and related to ‘jas’ affixes will optionally not be considered compounds. Three terms, ‘vibhasha’, anyatarasyam’ and ‘va’ are used u sed to indicate optionality. Panini and all the commentators have given us no indication that they are
supposed to be anything but synonyms. But the modern scholar Paul Kiparsky has wondered how could this be so, because Panini has vowed to eliminate every needless extraneous syllable and their must be a deeper reason to suggest the use of three different terms. Hence he has propounded propo unded the hypothesis in his wellargued study ‘Panini as a Variationist’ that the three terms ‘va’, ‘vibhasha’, ‘anyatarasyam’ refer respectively to three different kinds of options: those that are preferable(va), those that are marginal(vibhasha)and those that are simple options(anyatarasyam). If this is true than Kiparsky has shown us something which twenty-five hundred years of scholarship has overlooked. (8) atidesh sutras or extension rules: A rule is termed atidesha if it transfers certain qualities or operation to something for which they did not previously qualify. This is generally accomplished by the use of affixes ‘vat’/’mat’ ( like).Example: in rule 3.4.85 ‘loTo langvat’( the loT’s are like lang ) loT lakar affixes are being given g iven all the attributes unique to lang lakar affixes. Hence this is an atidesh sutra. (9) nipAtan or ad hoc rules: Certain nominal and verbal forms are underivable by any rules. Such forms, Panini accounts by just mentioning them. Such sutras are termed nipAtan. For example, in the domain of sutra ‘avayayIbhAva ’ (2.1.5), compounds (samas) of the type avayayi-bhava are being discussed. He discusses many rules which qualify a compound to be called avayayibhava. However, in Sutra: ’tishThadguprabhRtIni cha’ (2.1.17; all the elements of tishthadgu-group are to be considered avayayibhava)
he merely states that all the thirty-three elements of the ‘tishThadgu’ group listed in the Gana-Path (Catalog of nominals) are to be considered avayayibhava. Elements of set ‘tishThadgu’ are said to be avyayIbhAva by nipAtan. Example of Paninian Derivation: Siddhis of verb-form pathati( reads, singular pe rson, Present Tense ) When natives of Sanskrit uses sentences like ‘sah paThati’ ( he reads) they are, as all a ll the native speakers of a language, speaking unconsciously. Panini’s Grammatical Engine generates these correct forms from his rules, similar to how a Computer c ode using programmatic instructions generates desired output. Siddhi (Derivation) of paThati ‘paThan’, this is the verbal root as it is listed in Panini’s Dhatu-path and all forms of ‘paThan’ in speech are inflections of it.
1
2 3
4
‘ paThan paThan’
‘bhoovaadayo dhaatavah’ (1.3.1) This sutra establishes that ‘paTh’ is a verbal root. Only now, it is eligible for Verbal operations. ‘ paTh ‘updeshe ach anunasika ita’( 1.3.2 ) paTh’ A nasalized vowel (the ‘an’ of paThan’) is a marker and hence should disappear. ‘ paTh Dhatoh(3.1.91) paTh’ Since ‘paTh’ is a dhatu(verbal root) its operations come under the domain of this heading sutra. ‘Vartamaane Lat’(3.2.123) To indicate present tense, the ‘Lat’ affix should be used. ‘Pratyayah’, ‘parascha’( 3.1.1 and 3.1.2) An affix is always appended at the end of the stem. So this rule further makes a provision that affixes will be appended at the end and not the beginning. Hence the new form is ‘paTh’ + ‘lat’ and not ‘lat’ + paTh’ ‘ paTh paTh’ + ‘halantyam ’(1.3.3) ‘laT ’ The ending consonant is a marker. ‘tasya lopah (1.3.9) Markers are subject to elision. Hence the‘t’ of ‘lat’ disappears.
5
‘ paTh paTh’ + ‘lan’
‘updeshe ach anunasika it’(1.3.2) The ‘lat’ affix is actually ‘lant’ with a nasal ‘n’. The tradition of showing the nasal vowel( indicated by the sign of Chandra-bindu, a dot above the phonemes) had fallen into disuse, even by the time Vaman and Jayaditya wrote their commentary th called ‘Kasika’(6 Century C.E.) and no manuscript has been discovered to date with the nasal’s displayed. One good reason for this is that the nasal makes pronunciation so difficult. Besides its only role is as a marker which indicates elision. Hence, the ‘an’ of ‘lan’ disappears by 1.3.2.
6
‘ paTh paTh’ + ‘l’
‘lasya’( 3.4.77) ’tip-tas-jhi-sip-thas-tha-mib-vas-mas-ta-AtAm-jha-thAs-athAm-dhvam-iD-vahimahing’(3.4.78) ‘lasya’ above is meaninglessly in the genitive case. Hence the paribhasha ‘shashthi sthAneyogA’ (1.1.49) has to be invoked. According to it, items in the genitive case indicate ‘substituends’, hence ‘l’ is a substituend. The ‘l’ is replaced by any one of 18 affixes enlisted in the subsequent sutra ‘tip-tas jhi-…-vahi-mahing’(3.4.78) depending on whether the root has to indicate (a). atmanepada or parsmaipada endings (b) single, dual or plural number (c) first, middle or third person. ( Permutations: 3*3*2=18, hence the 18 affixes enlisted enlisted in Rule 3.4.78)
Now the next step becomes: ‘paTh’ + l -> {tip|tas|jhi|sip|thas|tha|mib|vas|mas|ta|AtAm|jha|thAs|athAm|dhvam|iD|vahi|mahing’} And our task is to cho ose exactly one affix that will append appe nd to ‘paTh’ in place of ‘l’ such that the inflection will take a parasmaipada ending (which is the ending the given root is assigned in the Dhatu-path Catalog and also found in Native Speech) and denote third person and single number. ‘laH parasmaipadam’ (1.4.99) The substituends of ‘l’ are termed ‘parasmaidpada’. According to this Sutra all the 18 1 8 afixes – being substituends of l – are now parasmaidpada. However, subsequent sutra – which is a niyama sutra – restricts parasmaidpada term for the first nine only, by assigning the last nine the term atmanepada. ‘tangAnau-Atmanepadam’ (1.4.100) ‘tang’ –a pratyAhARa formed by the ‘ta’ of the tenth affix and the ‘ng’ of the eighteenth and last affix ‘mahing’ indicating the last nine affixes – and the affixes ‘shAnach’, ‘kAnach’ are termed atmanepadam. Since, ‘paTh’ is a parasmaipada pa rasmaipada root, it will take one of the first nine affixes only; and we can now eliminate the last nine affixes which are atmanepaada. Therefore, the next step now is reduced to, paTh’ + l -> {tip|tas|jhi|sip|thas|tha|mib|vas|mas} Now Sutra 1.4.101 is invoked ‘tingastrINItrINI pratham-madhyam-uttamAH (1.4.101) The triplets of ‘ting’ in both parasmaipada and atmanepada are termed prathama (third person-madhyam (middle person) -uttama (first person) respectively.
Thus, the new distribution can be illustrated by the following Table:
Person Prathma (third person) Madhyama (middle person) Uttama (first person)
Parasmai-pada Triplet tip, tip, tas, jhi sip, thas, tha
Atmanepada Triplet ta, AtAm, jha thAs, athAm, dhvam
mib, vas, mas
iD, vahi, mahing
Sutra 1.4.103 discusses usages which determine the choice of middle person
endings. If the word ‘you’ is co-referential to a verb or is implicitly stated, the six affixes categorized as madhyama can be used. Similarly, Sutra 1.4.105 states that if the word, ‘me’ is co-referential with a verb or its meaning is implicitly stated, the six affixes categorized as ‘uttama’ can be used. This sutra is followed by Sheshe prathamaH (1.4.108) In the rest of the cases, use ‘prathma’ Since, ‘paThati’ is not co-referential with ‘you’ or ‘me’ and neither of these two meanings are implicitly assumed, ‘paTh’ qualifies for prathama endings by 1.4.108. Thus, we have further narrowed our search our search for the correct substituend to: ‘paTh’ + l-> -> {tip|tas|jhi} ‘tani-eka-vacahna-dvi-vachana-bahu-vachanAni-ekashaH’ (1.4.101) Each one of the three members of the triplets mentioned before in 1.4.100 are termed eka-vachana(single number ), dvi-vachana(dual number) and bahuvachana(plural) respectively.
Thus, a new distribution pattern now emerges:
Person
Parasmai-pada Triplet Single Single Dual Dual Plural Plural
Atmanepada Triplet Single Single Double Double Plural Plural
Prathma (third person) Madhyama (middle person) Uttama (first person)
tip
tas
jhi
ta
AtAm
jha
sip
thas
tha
thAs
athAm
dhvam
mib
vas
mas
iD
vahi
mahing
Since, in ‘saH paThati’ , (He reads) there is only one person, ‘ paTh’ must take an affix denoting Single number. Amongst the three remaining affixes ‘tip/tas/jhi’ , only the first affix ‘tip is in the Column for Single Number. Hence, path will accept ‘tip’ to indicate third person, Single Number, and parasmaipada ending.
7
8
‘ paTh paTh’+ ‘tip’
‘ paTh paTh’ + ‘shap’+ ‘tip’ 9 ‘ paTh paTh’ + ‘sha’ + ‘ti’ 10 ‘ paTh paTh’ + ‘a’ + ‘ti’
‘ting shit sarvadhatukam ’(3.4.123) ‘tip’ affix is a member of the set ‘ting’. This rule – a samjna sutra – classifies ting as a ‘sarvadhatuka’. Hence, all operations on the sarvadhatuka class of roots will now operate on: “ ‘paTh’ + ‘tip’ ” ‘kartari shap ’ This rule, a direct result of the preceding rule, now mandates that the infix (vikaran) ‘shap’ must be introduced before the ‘sarvadhatuka’ affix ‘tip’. Repeating sutras in operation # 5, the marker ‘p’ in both ‘shap’ and ‘tip’ disappear
‘lashaku ataddhite’ (1.3.8) This rule identifies the ‘sh’ of ‘shap’ as a marker. He nce the phoneme ‘sh’ disappears. Finally, we are ready to merge the three to form paThati(reads) . It has been correctly derived using the Grammar Engine of Panini and corresponds exactly to the manner it is spoken by the natives
11 ‘ paThati paThati’ Reads.
And each and every one of all the millions of verbal and nominal inflections in the speech of a Sanskrit speaker must undergo a similar set of operations to yield the correct form. Thus, every spoken word can be generated precisely by rules making it comparable to the Modern Discipline of programming. It should be borne in mind that the finite numbers of rules of Panini’s Grammar-Engine are capable of generating infinite forms. But each and every form generated by his Engine must correspond to the spoken language (loka) exactly which Paniniyas consider the only valid proof (pramana) of correct speech. Sometimes, however this is not possible and one word in an a n infinite set does not yield to the patterns. Panini takes this into account and creates one unique rule for that one anomalous word. For example, rule anudattatangit Atmanepadam (1.3.12) states that verbal roots in the Dhatu-path (the catalog of all known verbal roots which are around 2000 and is ancillary to A.) that are marked by the sign indicating low-pitch(anudatta) or have the marker (ng) appended to them take atmanepada endings. The verbal root ‘ji’ is not marked with either anu datta or ‘ng’ in the Dhatu-path and in regular usage does not have atmanepada endings (it takes parasmaidpada endings). However, native speakers in speech conjugate atmanepada endings if ‘ji’ is preceded by preverbs ‘vi’ and ‘par’. Panini took this irregular behaviour into account and coded rule (1.3.19). Thus, Panini didn’t allow one exception to be missed out viparAbhyAm jeH (1.3.19). from his comprehensive system.
Appreciating Panini’s Task in Codifying a Complicated Language like Sanskrit In English, if I were to make a sentence like ‘John’s book’, it will not make any difference in the structure of the sentence if I were to say instead ‘Mary’s book’, ‘Lizzy’s book’, and ‘Antonio’s book’. However, in Sanskrit the case-ending will differ according to the terminal phoneme of the nominal stem, as also by its gender and number (singular/plural). Some examples of this diversity are: Sanskrit Statement
TP -> APICR*
English Equivalent
sya pustakam Rama sya pustakam
‘a’ -> ‘sya’
Rama’s book
Gender; Number; Ending Phoneme; Masc.; Sing; ’a’
Seetaa yaah pustakam
‘aa’ -> ‘yaah’
Seetaa’s book
Fem; Sing.; ’aa’
eh pustakam Har eh pustakam
‘i’ -> ‘eh’
Harm’s book
Masc.; Sing; ‘i’
ayuh pustakam Pat ayuh
‘i’ -> ‘ayuh’
Husband’s book
Masc. ; Sing; ‘i’;
Gur oh pustakam oh pustakam
‘u’ -> ‘oh’
Masc.; Sing; ‘u’
Madhunah pustakam
‘u’ -> ‘unah’
Guru(Teacher)’s book Madhu’s book
Rajnah pustakam
‘an’ -> ‘nah’
Masc. ; Sing, ‘n’
tasya pustakam
‘t’ -> ‘sya’
Rajan’s ( King’s) book His ( tat ) book
Fem.; Sing; ‘u’
Masc. ; Sing.; ‘t’
*TP -> APICR = Terminal Phoneme - > Altered Phoneme Indicating Case Relationship Please note: the paradigms shown above are far from being a complete listing of the known inflections of the genitive case in Sanskrit. The actual number of distinct examples may run into more than one hundred forms all of which indicate the idea of someone possessing possessing something, something, in our case, case, a book.
Panini provides only one affix for indicating a singular genitive case relationship: ‘nas’ (Jas in HK Scheme of Transliteration). Therefore, it is Panini’s job to formulate a number of rules that will account for all the ‘x’ number nu mber of inflections, the native speakers uses unconsciously. In order to achieve this, he must first have an exhaustive listing of each and every known inflection– in our case the singular genitive – of the Nominal Stems. He has to study this list and discover the underlying phonological and morphological patterns. Once he has discovered the patterns, he creates rules. Every exceptional case (apavaada) has to be accounted for.
But ‘nas’ is only one of twenty-one affixes introduced in rule rule ‘su-au-jas-am-autshas….’(4.1.2) which govern the Sanskrit case-endings. All twenty one of these combine with nominal stems and each one of them generates hundreds of variant case-ending phoneme-forms as in the illustrated example of the affix ‘nas’ (Jas). (Jas). The case of the verbal roots and affixes is equally complicated. . Consider the following in English, (1) eat -> eats , (2) gain -> gains, (3) offer -> offers, (4) drink -> drinks, (5) read -> reads Conjugating a verb to imply present Tense is as simple as appending an ‘s’. Now look at equivalent forms in Sanskrit. (1) ‘at’ -> At ti* (eats), ti* (eats), ate(gains), (2) ‘labh’ -> labh ate(gains), oti(offers [an oblation]), (3) ‘hu’ -> ju -> juh h oti(offers (4) ‘pA’ -> pibati pibati (drinks). (drinks). (5) ‘paTh’ -> paTh ati(reads) ati(reads) *bold and italicized text indicates changes due to conjugation
Unlike English, not one of them terminates analogously. We have endings such as ‘ti’, ‘ate’, ‘ju..oti’, ‘bati’, ‘ati’ . One reason for divergent forms is that the v erbs belong to a different gana or group. The Dhatu-Path lists about 1967 verb roots (2014 including kaNDvAdi roots) divided into 10 conjugation classes (gaNas) to undergo peculiar operations. Each of these ganas are classified on the basis of distinct augments(‘shap’, ‘shyan’, ‘shnuh’, ‘shah’, ‘shnam’ , zero(luk) etc ) that tend to append to the verbal bases resulting in distinct grammatical operations. These verbal roots interact with the postverbal affixes (the set of ‘tings’) which are eighteen in number. These eighteen affixes get modified in ten ways( 10 * 18 = 180 ) ways to yield various tenses and moods, such as past tense ( ‘lan’-lakaar ), historical past tense( ‘liT’ lakaar ), ), present tense( laT lakAr’), imperative mood ( ‘loT’ lakaar ). ). For example: paTh* + shap** + tip = paThati; ‘(He) is reading; Present tense paTh + shap + Nal (modification of tip) = papaTh; He read; Historical Past (for historical characters like ‘Lord Buddha read’) + suT = paThet ; (You should) read; Injunctive Mood paTh + shap + yaasuT + *Verbal Root placed in the gana ‘Bhu-adi’. ** shap, the infix which appends exclusively to the ‘Bhu-adi’ group tan* + u** + tip = tanoti ; ‘spreads’: Present Tense tan + u + shya + tip= tanishyati; ‘will spread’; Future Tense
*Verbal Root placed in the gana ‘Tanu-adi’ gana **u is the unique infix for members of the ‘Tanu-adi’ gana hu* + shlu** + tip = juhoti; ‘(He) is offering (oblations); Present Tense, Singular, Third Person hu + shlu + jhi = juhvati; (They) are offering (oblations); Present Tense, Plural hu + shlu + mip = juhomi; (I) am offering (oblations); Present Tense, Plural, First Person hu + shlu + tip = juhotu; (They) should offering (oblations); Imperative Mood, Third Person
*Verbal Root placed in the gana ‘juhoti-adi’ gana **shlu is the unique infix for members of the ‘juhoti-adi’ gana and tend to ellision e llision Summary of the Major Topics of Ashtadhayi Sanjna or Terms: Panini has coined several technical terms in his A. which are sometimes meaningful (like samhita, samprasaran, anunasika) and sometimes meaningless (like ‘ghu’, ‘Ti’, ‘bha’). The former are generally considered to be b e pre-Paninian and the latter the creation c reation of the Sage himself which reflect his avowed concern for economy and are usually no longer than one syllable. He sometimes defines them; sometimes uses them without explicitly stating their meaning - implying that the term was in current usage and widely known. kno wn. And if the term refers to sets of grammatical items he merely lists the members that constitute that finite set indicated by the Sanjna. For example, ‘kta-ktavatu niShTha’ (1.1.25) enumerates the two constituents of finite set ‘niShTha ’. niShTha: niShTha: = {kta , ktavatu}
‘vRddhirADaich’ (1.1.1) phonemes ‘A’ ‘ai’ and ‘au’ are termed Vrddhi. vRddhi: = {‘A’, ‘ai’, ‘au’} Panini has used more than one-hundred o ne-hundred technical terms and the first two quarters of Chapter 1 concentrate on Sanjnakaran or Terminology-coining. Sandhi (Euphonic Combination): In English, a statement like ‘Who is going?’ in the speech of some native speakers speak ers often becomes ‘Who’s going’. In Sanskrit, this tendency for phonemes to fuse when in close proximity is extravagantly copious. Their can hardly be a sentence where at least one instance of Sandhi is not observable, if not more. Sandhis – given by Panini’s Sutra ‘paraH sannikarshaH samhita’ (1.4.108) – are morphophonological alterations that occur between terminal and initial phonemes of juxtaposed words (external Sandhi) or between morphemes within words (internal Sandhi). For example, in the sentence aha m A m Agachhami (I am coming), the terminal
phoneme ‘m’ of ‘aham’ (I) fuses with the initial phoneme ‘A’ of Agachchhami Agachchhami (am coming) to become ‘ahamAgachchhami ’. Sandhis have been popularly p opularly classified into three types: 1. Vowel Sandhi: between two vowels. For example: (Come then) becomes A A A Agachhet (Come tad = tad tad A + A + A Agachhet = Agachhet ( by the rules ‘akaH savarNe dIrghaH (6.1.101) (6.1.101) ) the two phonemes ‘A’ and ‘A’ fuse to become one phoneme ‘A’ 2. Consonant Sandhi: between consonants or between a vowel and a consonant. For example: t + Shivah = ta chchhivah Ta t + phoneme ’t’ transforms into ‘ch’ and phoneme ‘sh’ transforms into ‘chh’ 3. Visarga Sandhi: between Visargas (roughly like the ‘h’ sound in English) and other phonemes. For Example in Harih avadat (Hari spoke) avadat (Hari h + a avadat = ravadat = Hari r Hari h + The Visarga sound ‘h’ is transformed into the phoneme ‘r’. Sandhi can be used to make long strings of compound words and many poets have exploited this innate potential of their language. Banabhatta, the celebrated court-poet of th King Harshavardhana of Kannauj in his 7 Century Sanskrit Novel ‘Kadambari’ wrote several pages which consist of what is technically one single word using the principles of Sandhi and Compounding. Compound Words (Samaas) English has numerous compound words like class-room, foot-ball, frying-pan etc. These have chiefly evolved out of human-usage and are not the conscious coining of linguists. However, Compounding is peculiar to the idiom of Sanskrit usage. Writers freely use compound words of their own coining using Panini’s Compounding rules without any fear of unintelligibility. Excesses of usage for rhetorical purposes are as frequent as with Sandhi. Compounds are classified into the following four types: (6) avyayibhava(first term dominates in the final meaning of the compound) (7) tatpuruSha(the last term dominates in the overall meaning of the compound) (8) bahuvrIhi(neither (8) bahuvrIhi(neither dominate) (9) dvandva(meaning of both terms dominates)
DVANDVÁ or COPULATIVE
ITARETARA SAMAAHAARA EKASHESA
TATPURUSA (Inflectional) TATPURUSA or DETERMINATIVE NAN (Negative) SANSKRIT KARMADHAARAYA COMPOUNDS (Appositional)
Accusative Instrumental Dative Ablative Genitive Locative Dvigu
PRAADI (Prepositional, 1st kind) GATI (Prepositional, 2nd kind) UPAPADA (Compounds containing a upapada) SAMAANAADHIKARANA (The members --generally two-- are in apposition to one another)
BAHUVRIIHI or ATTRIBUTIVE VYADHIKARANA
(The members --generally two-- are not in apposition to one another)
AVYAYIIBHAAVA or ADVERBIAL . *from http://www.sanskrit-sanscrit http://www.sanskrit-sanscrito.com.ar/english/ o.com.ar/english/sanskrit/sanskri sanskrit/sanskrit8intro.html t8intro.html
Affixation: Discoveries of Sanskrit affixes and their meanings have be en instrumental in enhancing the deeper knowledge of Sanskrit words and have made possible neologisms using the principles established by P. Panini has described three hundred and seventy-five affixes of which ninety percent are those classed in krT and taddhita categories
Sanskrit Affixes
kRt
Others
taddhita
kRtya
tadrAja
**Ardha
nishThA
vibhakti
***sArva
sat
gha
Feminine
* sh-marked
Others
* sh-marked sArvadhAtuka and ArdhadhAtuka Affixes, includes the Unadi group. ArdhadhAtuka krt Affixes, ** ArdhadhAtuka affixes which include the Dhatu-affixes *** sArvadhAtuka which includes tings and the sh-marked vikaranas(augments)
The major group of affixes are discussed below: (1) kRt or primary Affixes: They serve to form derived nominal bases and generally denote objects of acts or abstract acts ac ts themselves. These are mostly appended to verbal roots. There are nearly one hundred and twenty-two Krit affixes described by Panini. In the chart above I have divided them into four groups: (a) kRtya: The seven affixes ’tavyat’, ‘tavya’, ‘anIyar’, ‘yat’. ‘kyap’, ‘Nyat’ and ‘kelimar’ are kRtyas and are used in the sense of ‘should be done’. Example: kR + tavya = kartavya; must do. (b) nishTha : kta and ktavatu are the only two elements of this group (c) sat: shatR and shAnach are the two elements of this group (d) shit sArvadhAtuk and ArdhadhAtuk affixes: several in number and includes the Unadis which are extremely irregular in behaviour. (2) taddhita or Secondary affixes: They also serve to form derived nominal bases but are appended to substantives to change the meaning of the nominal in different ways. 1100 rules in the A., covering most of the fourth and fifth chapter deal with taddhita affixes. The total number of taddhita affixes is two-hundred and seventeen. Examples are aN, chha, atsuch, kha. etc. (3) Feminine. Converts a given word from the masculine to the feminine gender. Examples: TAp, DAp, chap etc.
(4) Ting: These eighteen affixes append to verbal roots and are responsible for verbal inflections. Example, paTh + si = paThasi (you read). (5) Sup: These twenty-one affixes append to substantives and are responsible for Nominal Inflections. Example RAma + TA = RAmeNa (done by Rama). (6) DhAtu: These affixes form derivative verbs from primitive roots. For example: "kR" (to do -- a primitive root) can conjugate into ‘ karoti ’(does). If you want to express "a desire to do", Dhatu affix ‘san’ is appended and yields ‘ chikIrSati ’. ( He wants to do, he desires to do). Other examples are ‘ Nich’, Nich’, ‘yang, ‘yang, ‘san’ etc Karakas: Karakas are the equivalent of case relationships and are syntacto-semantic in Panini’s system. There are six karakas in Sanskrit. These are: 1. kartA Karaka (corresponds to the nominative case) 2. karma Karaka (accusative case) 3. karan Karaka (instrumental case) 4. sampradAna Karaka (dative case) 5. apAdAna Karaka (ablative case) 6. adhikaraN Karaka (locative case) The genitive(sambaandh) case is not recognized as a Karaka because it does not directly participate in the action denoted by the sentence. Post-Paninian Scholarship The study of Panini’s Grammar had spawned a vast commentarial literature in the past. The sages Katyayana and Patanjali have been mentioned previously. Katyayana intended to improve upon some of Panini’s rules where he felt they were inadequate. Inadequacies could either be due to changes that had crept into Sanskrit over the period of time between Katyayana and Panini or due to the oversights of Panini himself. Another purpose was to provide with fresh linguistic insights. insights. He has merely commented upon 1500 Paninian sutras in about 4000 vartikas(scholia) which are mostly one-liners and occasionally in the shloka meter(shloka-vartika). These vartikas are not an independent treatise and are only found in the commentary of Patanjali, the Mahabhashya, regarded as the second most important grammatical text after the Ashtadhyayi. Patanjali has commented upon 1701 170 1 sutras in 85 chapters called day-sessions (ahnikas). The commentary is in the form of a discourse between a student, a teacher’s aide and a teacher. The Mahabhashya is an extremely elegantly written treatise, composed in a very simple yet subtle style; and full of charming anecd otes and maxims. It is unique in being unpedantic which unfortunately is a rarity in S hastric texts.
Prominent names besides the three sages are Bhartrihari Bhartrihari(Vakyapadiya ), Vamana Jayaditya(Authors of Kashika), Kaiyata (‘Pradeep’ Commentary on Mahabhashya Mahabhashya), Bhattoji Dikshit Dikshit ( author of Siddhanta Kaumudi ), Kaunda Bhatt (Vyakarana Bhushan Sara) Nagesh( Voluminous author on Sanskrit Grammar, chief amongst them LaghushabdenduS LaghushabdenduShekhar hekhar , Paribahshendu-Shekhar, Udyot Commentary ). Special mention must must be made of the Siddhanta Kaumudi of Bhattoji Dikshit written in the seventeenth century by a learned Brahmin of Maharashtra revolutionized the way Grammar was taught in the length and a nd breadth of the entire subcontinent. Following F ollowing the ‘Prakriya’ School which emphasizes derivation of correct forms above others and teaches the A. without regard to its spatial order; it beca me the most important pedagogical text and replaced traditional teaching of Grammar through A. Many modern scholars like Ram Nath Sharma and the Arya Samajis likes Brahma Dutt Jijnasu and Yuddhishthir Mimansaka consider the SK’s prakriya system unscientific even unfortunate. The latter have leashed a diatribe of vehemence on its methods and equated th the downfall of Indian Culture with the emergence of the Prakriya Schools in the 11 Century. The use of mythological paradigms by Bhattoji to illustrate rules has done much to increase its popularity amongst teachers who co nsider it a text with a dual aim: a im: teaching of grammar and inculcating religiosity. In modern times, some of the world’s most brilliant linguists - both Indian an d Western have studied Panini and produced a plethora of novel research. Paniniyas such as George Cardona, Madhav Deshpande, Ram Nath Sharma, S.D. Joshi, Roodenberg, Devashthale etc have written very precise treatises and laid bare all the intricacies of the technique in English so that a modern student has no longer to depend on his teacher or a cryptic Sanskrit commentary and can comprehend the intricacies of Sanskrit Grammar from his reading desk. The research P. has inspired is extremely meticulous, scientific and detailoriented. Scholars have written lengthy books to discuss the use and significance of single words like ‘cha’ (and), the markers (‘The Anubabdha s of Panini’), the optionalitysignifying words( va, vibahsa , anyatarasyam )( ‘Panini as a Variationist’) etc. There is nothing new in this as from the times of Patanjali itself, exhaustive discussion on single sutras (example Samarthanhnika: single chapter on Sutra 2.1.1 ‘samarthah padavidh samarthah padavidhi’ i’ ) was the norm rather than the exception. e xception. Modern Civilization with its rapid means of communications and transmittal of knowledge has catalyzed this process. Summing up my paper I would wou ld like to mention the A. is not only the most important means for a proper understanding of the Sanskrit Language and Grammar, its techniqu e is so interesting in itself that one can study it for the mere pleasure of learning a very ve ry engaging and intellectual system which American Linguist Leonard Bloomfield felt was ‘ one one of the greatest greatest monuments monuments of of human intelligence intelligence ’ (quoted in ‘Panini: A Survey of Research, George Cardona, pg 243).
Bibliography Sharma, Ram Nath The Ashtadhyayi of Panini, Volume I,II and an d III: Introduction to the Ashtadhyayi as a Grammatical Device. Delhi: Munshiram Manoharlal Publishers. Brahmadatta Jijnasu Sanskrit paThan-pAThan ki anubhut saraltama vidhi, part I (in Hindi). Sonepat: Ramlal Kapur Trust Cardona, George Panini: His Work and It’s Traditions, Vol. I. Delhi: Motilal Banarsidas Panini, A Survey of Research. Delhi: Motilal Banarsidas. Recent Research in Paninian Studies. S tudies. Delhi: Munshiram Manoharlal. Joshi, S.D. Patanjali’s Vyakarana Mahabhashya: Paspashahnika. Poona: University of Poona Kiprasky, Paul Panini as a Variationist. Poona: University of Poona/Cambridge MIT P ress Abhayankar, Kashinath Vasudeva and Shukla, J.M A Dictionary of Sanskrit Grammar: Baroda: Oriental Institute. Devasthali, G.V. SaraSiddhantaKaumidi of Varadaraja: Poona: University of Poona The Anubandhas of Panini: Poona: University of Poona Staal, Frits Mantras and Rituals: Rules Without Meaning: Delhi: Motilal Banarsidas Mahulkar, DD (Editor) Essays on Panini: Shimla: Indian Institute of Advanced Studies Allen, W. Sidney Sandhi: The Theoretical, Phonetic and Historical Bases of Word Junction in Sanskrit: Hague: Montana and Co.
